How Much Pipeline Coverage Do You Need?

Is it 3x? Maybe! Let's do some math.

Jan 16, 2025

tl;dr - Determine your pipeline coverage goal empirically when possible, but you can estimate it with some math:

The 3x “rule”: Nobody knows where it came from and it’s certainly not one-size-fits-all.
Coverage benchmarks: Based on benchmarks from Benchmarkit, companies with smaller ACVs carry less start-of-quarter pipeline coverage, likely due to shorter sales cycles. Companies with $25k-$50k ACVs seem to be playing a slightly riskier game.
Determine your target: If there’s not enough empirical data, use an equation from Insight Partners and some basic assumptions to arrive at a reasonable estimate.

Read on for the full story.

Welcome back to Pipeline Month at Uncharted Territory. Last week I wrote all about pipeline coverage: what it’s really telling us, how to measure it (correctly) and how to use it.

I punted on one little detail. Once you calculate your pipeline coverage, how do you know if you should be comfortable or terrified about hitting plan? To get there we need a way to figure out what our pipeline coverage ratio ought to be.

Simple right? 3x. (Or not.)

For those of you who don’t know, sales leaders have usually felt safe having 3x pipeline coverage at the start of a period. Why? Well, no one actually knows. I’ll defer to the godfather of pipeline metrics, Dave Kellogg, for some context:

So where does the magical 3x coverage ratio come from? I don’t know the history, but I can say that long before I saw — and I mean years — my first salesforce automation system, I heard sales managers speak of the rule of three. It makes sense: 2x seems tight and 4x seems rich. So, through the Goldilocks Principle, we ended up with 3x.
— Dave Kellogg writing on Kellblog in 2013

He wrote that in 2013—nearly 12 years ago. Even then, the 3x “rule” was ancient received wisdom. It seems to have arisen from the primordial soup of sales management that existed *gasp* even before the CRM.

Maybe our elders are wise and it really should be 3x. But why? Well, we know pipeline coverage isn’t the inverse of win rate so we can’t just plug in 1 over our win rate and be done.1

So how much pipeline do we really need to feel comfortable at the start of a period? (Or at least less terrified.) Let’s dig in.

Benchmarks

Before we figure out the right pipeline coverage ratio for your team, let’s take a look at some benchmarks. They won’t tell tell us everything, but at least we’ll know what kind of company we’re in.

A couple weeks ago I put out a call for useful pipeline coverage metrics on LinkedIn and heard back from Ray Rike, CEO of Benchmarkit (and one half of The Metrics Brothers).

Ray does benchmarking for a living. Thanks to him, we have at least a partial answer to this question from a recent benchmarking exercise he ran on quarterly pipeline coverage.2 See below.

Even this chart reveals bias towards the 3x rule—the column groups are centered on 3-3.5x coverage while anything below 2.5 and above 4 gets clustered together. There might be some interesting spread in the underlying data but we can’t see that.

I don’t want to over-interpret this one chart, but a few things do stand out:

The big ol’ 50% column in the 3x-3.5x group belonging to $10k-$25k ACV cohort seems a little artificial. I suspect some folks have turned the 3x rule into a self-fulfilling prophecy.
The $10k-$25k cohort doesn’t have anything above 3.5x coverage. This is probably because they have shorter sales cycles with more create-and-close pipeline that will originate within a quarter. Having 4x pipeline coverage would be going overboard and mean overspending relative to plan.
The $50-$100k group seems to be the most conservative. They greatly exceed the population numbers for the 3.5x-4x and > 4x groups.
The 38% of the of the $25k-$50k cohort that are running < 2.5x coverage seem like they’re playing a risky game relative to the rest of the population. While I don’t have great benchmark data on sales cycle length by ACV, it’s plenty common for these deals to take more than 90 days to close. This implies relatively little create-and-close in-quarter as well as a decent likelihood of pushes into next quarter.

One terrifyingly important thing this chart doesn’t tell us? Whether these teams ultimately hit their number. All we can really glean from this is where they stand, not whether these are the optimal ratios to succeed at hitting plan.

So what should my pipeline coverage ratio be?

Let’s get one thing out of the way. The best way to figure out what your pipeline coverage ratio should be is to already have lots of high quality historical data. If you have 20+ quarters of coverage snapshots and subsequent performance across a pretty stable business, then you can just calculate what you need empirically.

If you’re not swimming in high quality data (and many of us aren’t), you’ll need to make some big ol’ assumptions. However, as we’ll see, those assumptions plus some reasonable math can guide you to a good starting point that’s better than just 3x.

I’m going to lean on an equation provided by Jeremey Donovan at Insight Partners. See the right side of the slide below (note that the chart on the left side is just an example, not a recommendation).

Courtesy of Jeremey Donovan at Insight Partners

It’s important to note that all of these terms are expressed in dollars—either as percentages of pipeline or actual dollar amounts. As we’ll see, this can make things tricky in a few ways.

The other big challenge is you need some data points that require either a) some assumptions or b) historical data. Thankfully the particular historical data necessary may be available to you even if you haven’t kept good historical snapshots of pipeline.

Here’s the breakdown of the terms and some words of caution:

Create-and-close (c) - These are the dollars that don’t exist at the start of the period that you think you’re going to close by the end of the period, expressed as a percentage of total bookings. That’s a bit tricky. For a commercial team operating at a high velocity and a short sales cycle, you could estimate this for a quarter based on your normal monthly opportunity creation rate, ACV and sales cycle. For example, if your sales cycle is 30 days and you create, say, 20 opps per month, you could assume the 40 opportunities created in the first two months of the quarter could contribute to your overall coverage. To get a percentage you’d need to consider that this is likely to represent 66% of your overall bookings because you’d start the quarter with the 20 opps created in the final 30 days of the previous quarter. Note that this is particularly sensitive to seasonality. c may be *very* different in calendar Q3 than in Q2.
Win Rate (w) - This win rate is defined in dollar terms, not deal terms. In my experience most of us quote win rates as the percentage of closed deals won (e.g. we closed 100 opps and won 30 of those = 30% win rate). The dollar win rate might be quite different if there’s a meaningful variation in deal sizes.
Push Rate (p) - The % of starting pipeline that typically pushes. Notionally this data exists by default in Salesforce since opportunities track close date change history. You (or your ops team) should be able to go figure out how many opps typically push out of quarter. However, it can be tricky to back into these pushes as a percentage of starting pipeline3.

For completeness, let’s walk through the full quarterly calculation example for a team with an ACV of $20k, a 45-day sales cycle and a 20% win rate.

Create-and-close (c) - We can estimate we’d have a chance of closing half of the pipeline created during the quarter (everything created by day 45 in a 90-day quarter). Assuming steady state opportunity creation and that 45-day sales cycle, we’d likely enter the quarter with about half of the pipeline created last quarter still in play to create bookings. So we can set c in this case to 50% because about half of our bookings will come from the pipeline created during the current quarter and half will come from the previous quarter.4
Win rate (w) - Because we’re making a radically simplified assumption that pretty much every deal ends up at $20k, we can say that this number is equivalent to our 20% win rate in both deal terms and dollar terms.
Push Rate (p) - In this case I’ll assume that the push rate is essentially 0. This term is a percentage of the starting pipeline that pushes. Since we’re looking at a 45 day sales cycle and quarterly pipeline coverage, there shouldn’t be much that pushes into the next quarter.

Putting it all together, here’s how the equation would come out:

$r = \frac{1 - 0.5}{0.2 * (1 - 0)}$

$r = 2.5$

Based on the myriad assumptions above, you’d need a pipeline coverage ratio (r) of 2.5x.

We made a lot of simplifying assumptions to get here. However, when we look back at the benchmarks in the earlier section it tracks reasonably well. Half of the $10k-$25k benchmark cohort maintained a ratio less than 3x.

So, are you comfortable or terrified if this is your team? If your target was $1M, and you entered the quarter with $2.5M in pipeline you could be semi-comfortable. (Or at least not terrified.)

Share all this math with a colleague.

A brief coda

Ok, this got quite math-y and I resisted making it even more so. However, I’m pretty convinced a better way to think about this in the absence of lots of historical data is to get more statistical with it.

In particular, the simplifying assumption of dollar win rate being equivalent to deal win rate is a problem. Win rate, in particular, is actually binary: you win or you lose. This is why dollar-weighted win rates can be problematic. Whether you win or lose that $50k deal is binary—you either get the $50k or you don’t. Just because your overall win rate is 25% doesn’t mean you get to keep $12.5k on a lost deal.

We also know that an average contract value masks large variations. We’ve all seen deals jump in value or (sadly) suddenly decline after wrestling with a CFO. The right way to think about contract values isn’t as a single value but more as a probability distribution centered around a certain value.

Finally, sales cycles are the same way. An average of 45 days (even a median) masks some big differences. Some deals will take 90 days, some will take 10.

I can’t help but think that if we don’t have much empirical data, we’d be better off running an analysis like Nate Silver does for elections—build a model with some different probabilities and run it thousands of times to see the most likely outcomes.

If anyone’s interested in this type of simulation-based approach let me know. I might just make it a little weekend project at some point in the future.

That’s good, because 3x implies a 33% win rate, which not everyone has. A 20% win rate would imply you need 5x coverage which gets very expensive very fast.

If you’re wondering, Ray shared the overall measurement definition: “Pipeline Coverage Ratio was collected for most recent 3 months (quarter) and for trailing 12 months.” The full data are not published yet so I’m not sure of all the details like sample size, etc. However, Ray does this for a living so I’m sure it’s high quality.

A good analyst could get you there with Salesforce OpportunityHistory, some code, a lot of elbow grease and even more coffee but it’ll be painful.

Astute readers will note that a very rough calculation for create-and-close when the period is a quarter (~90 days) and the sales cycle is < 90 days is (90 - sales cycle) / 90.